• Please review our updated Terms and Rules here

486 to 386 problem

Sounds like it might be a divide-by-two circuit, in which case, you'll want an 80MHz crystal.
<rummages around...digs thru pile of dead parts>...Aha! Here 'tiz...
Ok, FWIW, I'm looking right at an old board here, and it has a 40MHz 386 DX chip installed, and the crystal is a 66MHz. I've never seen this board running, so I can't swear to what speed it actually ran at. As always, YMMV.

--T
 
Last edited:
Yeah, that makes sense. The original was a 25MHz crystal and a 486 DX/2, which actually 'ran at' 50MHz, correct? The 486 DX/2 is a 'clock-doubler', so on a 25MHz clock it would (apparently) 'run at' 50MHz. Going back to a 386 chip makes it run at half the speed of the oscillator. (And the board I have prob'ly only went up to 33MHz, even though the CPU is rated for 40).

--T
 
Actually, the original processor was a i486SX-25, I just tested the 486DX2-66 to see what would happen. I do see what you are saying thou.
 
CheckIt results show that the computer is indeed running at 20 MHz.
Terry Yager said:
Going back to a 386 chip makes it run at half the speed of the oscillator. (And the board I have prob'ly only went up to 33MHz, even though the CPU is rated for 40).

--T
Click to expand...
So your saying If I replace the crystal with a 80 MHz (if I can find one), that will fix the problem?
 
Best guess, mebbe. I have seen 80MHz crystals on boards of that vintage before. Then again, there's mine, which I mentioned earlier in this thread. It's *s'pozedta* be able to run at 40MHz, but it's not reliable at that speed, so no guarantees.

--T
 
Last edited:
Crystal

Crystal

I have located a 386DX-40 machine. It has a socketed 80Mhz crystal (secured with a cable tie).
The motherboard is an Elitegroup UM386 - same manufacturer and from around the same time.
Based on that info, I think that there's a very good chance that an 80Mhz crystal is your answer.
 
Well, I know what I have to do; the problem is where do I find the 80MHz crystal? There is one on ebay, but he's selling them in quantities of 20. And the store where I knew I could find the crystal at, closed..
 
The crystal on my EliteGroup motherboard has a part number of: KDN-110S 80.000 Mhz
The crystal on my other 386DX-40 motherboard has a part number of: KDN-020C 80.000 Mhz

They look physically the same to me - perhaps an electrical difference.
Google with "KDN-110S" and "KDN-020C" shows lots of hits.

As the your EliteGroup 3000 motherboard is 'closest' to my EliteGroup motherboard, I'd put preference in the 'KDN-110S 80.000 Mhz' version.
 
modem7 said:
They look physically the same to me - perhaps an electrical difference.
Google with "KDN-110S" and "KDN-020C" shows lots of hits..
Click to expand...


My educated guess...perhaps one is for a series tank circuit, the other for parallel. Or, a difference in the cut (X,Y) of the crystal.

Kent
 
When the 'tin can' has only two leads, the can only contains a crystal.
These 4 pin cans contain extra components to create a self-contained oscillator. You supply 5 volts (or whatever) across two pins and you get an oscillating signal out of the other two pins.
And so maybe the difference in part numbers is the supply voltage, or the accuracy of the output, ....
 
You must log in or register to reply here.
Back
Top