• Please review our updated Terms and Rules here

Turbo XT refuses to engage the turbo!

M

Mike Chambers

Veteran Member
Joined
Sep 2, 2006
Messages
2,621
I have this clone board, I've had it forever. I have never been able to get it to go into turbo mode! Topbench reports 4.77 mhz no matter what.

The jumper has no effect. Ctrl-alt-plus or -minus have no effect. I even tried manually toggling bit 2 of port 61h high. The bit goes high, verified by rereading, but no effect on speed!

Any ideas? I don't know the make and model, but this is the board:

McyI5vJ.jpg
 
What does that jumper just below the keyboard connector do? I see that it has two positions enable/disable and that it's set to "disable". And whose BIOS does it use?
 
That jumper enables/disables memory parity. As far as the BIOS, I'll check it tonight after work. I don't believe it says during the POST. I might have to peek at the ROM contents for a hidden string.
 
Unless my old eyes are deceiving me, from the photo, it looks like a D70108 = V20.

I can't make out an i8284 clock generator chip, however. Is that function part of the QFP?
 
Does the turbo LED illuminate at all when using the jumper or hot keys? A really simple way to check if the speed is changing is to type the DOS command ECHO (Ctrl-G) to make a beep. Usually on XT-class machines, the faster the CPU speed, the higher-pitched the PC speaker beep will be.
 
The BIOS is important if the turbo mode is engaged via keypress. In other words, the BIOS must contain the speed-switching code for this particular board.
 
Would it be possible to get a picture of the QFP clear enough to read the markings? If it’s a standard part with a datasheet that has anything to do with timing generation it might reveal what port to bang on.
 
vwestlife said:
Does the turbo LED illuminate at all when using the jumper or hot keys? A really simple way to check if the speed is changing is to type the DOS command ECHO (Ctrl-G) to make a beep. Usually on XT-class machines, the faster the CPU speed, the higher-pitched the PC speaker beep will be.
Click to expand...

I'll need to hook up an LED and see. I'm still not back home yet.
 
Eudimorphodon said:
Would it be possible to get a picture of the QFP clear enough to read the markings? If it’s a standard part with a datasheet that has anything to do with timing generation it might reveal what port to bang on.
Click to expand...

The image was, but the forum seems to have shrunk it a bit. Here's the original on Imgur: https://i.imgur.com/McyI5vJ.jpeg

It seems to want to rewrite the url if you click on it, at least for me, which doesn't let you make it large enough. You may have to copy and paste it manually into a fresh tab.

That's a wood splinter that fell onto it by the way, nothing metal!
 
Okay, so, that MT86G011 chip does appear to contain essentially all the XT standard chipset components, according to a mention in a manual for a “R.A.C.E.R. II” diagnostic card manual, but I failed utterly to find a datasheet that would reveal any technical details. According to the diagnostic manual it appeared in a computer called the “ACER 710”, maybe it’s remotely possible you could find a speed control DOS utility for that machine? If a keyboard sequence is necessary to switch it to fast mode it could be almost anything, I don’t remember them matching between any of the Turbo XTs I owned back in the day…
 
I found the same document, but I've been unsuccessful finding any utility for an Acer 710.

I had some time to rip the BIOS tonight, It's a Phoenix.

It actually does say so on post.

I threw the ripped BIOS into my emulator though. This is what it looks like, if it's familiar to anyone.

turbobios.png
 
Last edited:
Hard to say if the BIOS matches the motherboard. From appearances, it's a BIOS that someone burned into an EPROM sans any labels, so it might not support speed changing.
 
It very well could be that it's not the original BIOS actually, good point.

I suppose it's semi-moot as I've got another turbo XT board arriving in the mail today that I'll be able to use, but it still would be nice to figure this one out eventually.

I actually experimented with upgrading the 14.318 MHz crystal on this one last night. At 24 MHz, it started to POST, but the memory count never began. My memory probably wasn't working at all at that speed, so the BIOS only ran as far as it could before needing to use it.

At 20 MHz, the system ran fine, theoretically at a CPU clock of 6.67 MHz? However, benchmarking utilities still reported 4.77 MHz. Maybe I'm misunderstanding what replacing that crystal does. I thought it would increase the bus (f/4) and CPU (f/3) speeds when not in turbo mode. I know it would break CGA, but that doesn't bother me.

I put the original crystal back on since apparently, performance didn't improve.

I also have a "PCII System Board" turbo clone board that's refusing to POST at all, no matter what I do, though that would be a topic for another thread.
 
Last edited:
Simply replacing the crystal will always result in benchmarks reporting 4.77MHz, no matter what the actual speed is. When you replace the crystal, you're also making the timer run faster in lockstep with the CPU.
 
Ah, of course! That crystal controls everything. I knew it felt faster for sure, which is why the benchmark result confused me. That explains it.

I don't like the timer being affected, so I'll leave the crystal as stock.
 
That's one reason that the BIOS support is important on a lot of the "we'll just change the CPU clock" approaches. You need to change the time-of-day/"tick" scaling to conform; some routines, such as INT 13H floppy services will also drop the speed to 4.77 MHz temporarily. The alternative is to run the timer from the 4.77 MHz source, independent of the CPU. That requires a board designed for that.
 
That Flat-Pack IC should have all the hardware to do turbo the correct way; There’s a 30mhz crystal near the 14mhz one, I’m sure it is there to be divided by three for a 33/66% duty cycle 10mhz cpu clock when it’s enabled. (That fits the specs of the ACER machine the same chip was used in.) It’s all down to figuring out how to enable it. A wrong BIOS is as good a theory as any, I wonder how common that chip actually was in late-era XT clones.
 
Right, the 30 mhz is what I would have preferred to upgrade, but of course it's no good unless the turbo mode can be enabled.

I suppose I'll need to find the V20HL rated at 16 mhz if I really want to go wild with overclocking though. I wonder if those can be pushed closer to 20. I think it'd be fun to have a V20 machine that can give a 286 a run for it's money with VGA, Sound Blaster, etc. Something that can handle early 90's games well. At least the ones that didn't require protected mode yet! Ultima 6, Star Trek 25th anniversary game, stuff like that.

Faster memory would also probably be necessary unless I want to start inserting wait states, which I don't. ;)
 
You must log in or register to reply here.
Back
Top