Author Topic: OLD COMPUTER  (Read 3674 times)

0 Members and 1 Guest are viewing this topic.

Offline MrsR

  • Regular Contributor
  • *
  • Posts: 118
  • Country: au
OLD COMPUTER
« on: February 11, 2018, 03:48:09 pm »
Hi!
I have a 486DX 33 computer bought in 1990 and still used.
The hard drive packed it in, a 2.5 gig MAXTOR IDE HD,
I haven't played around with this computer in a long time (using it for CAD work ).
 
When I removed the HD I found I had put the jumper in for master but also put a jumper in across pin sets 2 and 3 in effect shorting them out
what  I am asking is why did I do it. I can't find any info on the net about it but there must have been a reason as it worked for 15 years
I have a 2.8 gig MAXTOR IDE HD on order and I need to know if I have to do the same to this.

It might sound a bit crazey me asking this but I haven't needed to work on any computer for years and I just forgot due to the length of time.

THANKS GUYS!!!!!
 
 

Offline soubitos

  • Frequent Contributor
  • **
  • Posts: 331
  • Country: gr
    • I sell on Tindie
Re: OLD COMPUTER
« Reply #1 on: February 11, 2018, 03:59:46 pm »
 

Offline picandmix

  • Frequent Contributor
  • **
  • Posts: 394
  • Country: gb
Re: OLD COMPUTER
« Reply #2 on: February 11, 2018, 04:13:01 pm »
As its 1990 its probably worth a lot now we have the Meltdown and Spectre bugs that affect all post 1995 chips !  ;D

 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1116
  • Country: ca
  • Hardcore Geek
Re: OLD COMPUTER
« Reply #3 on: February 11, 2018, 04:21:03 pm »
When I removed the HD I found I had put the jumper in for master but also put a jumper in across pin sets 2 and 3 in effect shorting them out

The old drive is probably an 82560?  The first jumper (J50, closest to the IDE connector) controls master/slave.  Jumper installed is Master/1-Drive.  Open is slave.  The next one, J48, is for Cable Select.  For CS, you leave J50 open and short J48, the drive will automatically CS master or slave depending on cable position (when the system supports CS, of course.)  The third one, J46, limits the drive to reporting only 4096 cylinders for systems that would have a BIOS hang with more than 4096.  The other positions are "Factory Reserved"

The jumpers going sideways is just the way of storing unused jumpers.  If you look at the board or check the resistance between the pins that the jumper is on, you'll find that they are already connected together, so storing the jumper there does nothing.  You can also store a jumper with one side hanging off into nowhere, in the space where the second pin on J44 would be.

They usually (unless specified otherwise in a bulk lot buy) shipped them with a jumper on J50 for MASTER and one spare jumper in case you needed to jumper for 4096 cylinders.  You never need more than the two jumpers, so they supplied two, just normally only one or none was actually required unless you needed CS or MASTER + 4096 CYL LIMIT.

Code: [Select]
                                        +---+---+---+---+---+
                                        |J50|J48|J46|J44|J42|
  +-------------------------------------+---+---+---+---+---+
  | Master/Slave                        |   |   |   |   |   |
  |  Only Drive in single drive system* | J |   |   |   |   |
  |  Master in a dual drive system*     | J |   |   |   |   |
  |  Slave in a dual drive system       | O |   |   |   |   |
  +-------------------------------------+---+---+---+---+---+
  | Cable Select                        |   |   |   |   |   |
  |  Disabled*                          |   | O |   |   |   |
  |  Enabled                            |   | J |   |   |   |
  +-------------------------------------+---+---+---+---+---+
  | 4092 Cylinder Option                |   |   |   |   |   |
  |  Disabled*                          |   |   | O |   |   |
  |  Enabled                            |   |   | J |   |   |
  +-------------------------------------+---+---+---+---+---+
  | Factory Reserved                    |   |   |   | O |   |
  +-------------------------------------+---+---+---+---+---+
  | Factory Reserved                    |   |   |   |   | O |
  +-------------------------------------+---+---+---+---+---+
    J=Jumpered O=Open *=Default

Your new one will just need to be set for MASTER.  What model is the replacement?

P.S. You probably could have just replaced it with a Compact Flash card or something.  :)
« Last Edit: February 11, 2018, 05:03:14 pm by drussell »
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1116
  • Country: ca
  • Hardcore Geek
Re: OLD COMPUTER
« Reply #4 on: February 11, 2018, 04:42:50 pm »
...
<image>
...



The diagram posted by soubitos is a bit misleading, since it is missing a pin in the CS position as indicated by the purple "?" above. J46 has two pins, as correctly shown in the M / S diagrams.  J44 is the only one missing a pin.  (At the "top".)

In any case, the main answer to your original question, "why did I do that?" is...  Because it is a convenient place to shove the extra jumper.  :)  It is not required.  You can just omit it if you like, save it in the junk box.  :) 

The one jumper for MASTER is all you need.
 

Offline Ampera

  • Super Contributor
  • ***
  • Posts: 2508
  • Country: us
    • Ampera's Forums
Re: OLD COMPUTER
« Reply #5 on: February 11, 2018, 09:12:30 pm »
These drives may not have cable select available. I know my 1996-1997 drives do not (my oldest drives). Cable Select came out a bit later.
C/C++/Java Programmer, Legacy hardware enthusiast, madman.
If it's broken, I probably did it.
EEVBlog IRC: irc.austnet.irc #eevblog
 

Offline imidis

  • Frequent Contributor
  • **
  • Posts: 427
  • Country: ca
Re: OLD COMPUTER
« Reply #6 on: February 11, 2018, 09:45:55 pm »
33 mhz? Did you never upgrade it to 75 mhz?  :'(

Yeah, most only needed one jumper so the other is usually in a 'storage spot'

For the yougins....

You set master/slave by setting the jumpers, IIRC windows liked to be the C: drive, so windows drive master second or cdrom etc slave. Or if it was ATA66 (IIRC) and you had a compatible cable (master connection = master drive, slave = slave drive) CS on the jumpers.

Also, jumpers were required for things like setting the clock/voltage of the CPU.

All this new fangled automagic bios stuffs now.
« Last Edit: February 11, 2018, 09:53:47 pm by imidis »
Gone for good
 

Offline Ampera

  • Super Contributor
  • ***
  • Posts: 2508
  • Country: us
    • Ampera's Forums
Re: OLD COMPUTER
« Reply #7 on: February 11, 2018, 10:58:43 pm »
I've built two manual jumper computers recently, and as a programmer of platforms nobody programs for anymore, I can say that the primary fixed disk drive is letter C, with the secondary (sometimes ATAPI)  being D. For computers with support for more than two fixed disks, the letters go up from there.

Also, in terms of upgrade, unless an overdrive chip is found, it's possible that machine can only go up to a 50 mhz upgrade (normally 40mhz if you have VLB cards). 486es, though, fun machines to work with.
C/C++/Java Programmer, Legacy hardware enthusiast, madman.
If it's broken, I probably did it.
EEVBlog IRC: irc.austnet.irc #eevblog
 

Offline imidis

  • Frequent Contributor
  • **
  • Posts: 427
  • Country: ca
Re: OLD COMPUTER
« Reply #8 on: February 12, 2018, 12:06:50 am »
I've built two manual jumper computers recently, and as a programmer of platforms nobody programs for anymore, I can say that the primary fixed disk drive is letter C, with the secondary (sometimes ATAPI)  being D. For computers with support for more than two fixed disks, the letters go up from there.

Also, in terms of upgrade, unless an overdrive chip is found, it's possible that machine can only go up to a 50 mhz upgrade (normally 40mhz if you have VLB cards). 486es, though, fun machines to work with.

Oh yeah, I think I did the overdrive on that one. Must have. I know I went from 33 to 75.

But apparently if I am reading this right (below) AMD and it is correct if you replace the voltage regulator it can be done...  :-\

https://en.wikipedia.org/wiki/Am5x86 (just looking at a chip I happened to have) Whoops mines the early socket 3 version.
« Last Edit: February 12, 2018, 12:37:55 am by imidis »
Gone for good
 

Online james_s

  • Super Contributor
  • ***
  • Posts: 9579
  • Country: us
Re: OLD COMPUTER
« Reply #9 on: February 12, 2018, 12:20:12 am »
You've been using the same 468-33 steadily since 1990? That's impressive in many ways. That's a classic machine, I remember well drooling over those back when I was far too poor to actually own one.
 

Offline Ampera

  • Super Contributor
  • ***
  • Posts: 2508
  • Country: us
    • Ampera's Forums
Re: OLD COMPUTER
« Reply #10 on: February 12, 2018, 01:18:37 am »
There are drop in socket to socket VRM adapters, but they are rare (I've never found one).

Some collectors do still readily use pre-DX2/DX4 486's. They are often some of the best chips available for pre-3D DOS games.
C/C++/Java Programmer, Legacy hardware enthusiast, madman.
If it's broken, I probably did it.
EEVBlog IRC: irc.austnet.irc #eevblog
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1116
  • Country: ca
  • Hardcore Geek
Re: OLD COMPUTER
« Reply #11 on: February 12, 2018, 06:29:36 am »
These drives may not have cable select available. I know my 1996-1997 drives do not (my oldest drives). Cable Select came out a bit later.

Most drives by 1996-97 had CS...  Pretty much everything that was in the multi-gig range had CS, though Western Digital was a bit slow to adopt it into their "EIDE" standard, etc. before ATA-2 came out.  It was a common manufacturer-specific extension before ATA-2.

As for the OP's drive, it is obviously much newer than the machine itself, the CPU, etc. but the 82560 that he probably has does indeed support CS.  Whatever original drive it had back in the day almost certainly did not support CS, nor will the machine, so it really doesn't matter.  They just needs to jumper for MASTER and will be all good to go.
 

Offline Whales

  • Frequent Contributor
  • **
  • Posts: 877
  • Country: au
    • Halestrom
Re: OLD COMPUTER
« Reply #12 on: February 12, 2018, 06:39:38 am »
Quote
I have a 486DX 33 computer bought in 1990 and still used.

Stop teasing us, we want pics  ;)  Software too

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1116
  • Country: ca
  • Hardcore Geek
Re: OLD COMPUTER
« Reply #13 on: February 12, 2018, 06:42:00 am »
33 mhz? Did you never upgrade it to 75 mhz?  :'(

75 MHz would be difficult on any early 33 MHz machine.  You would have to change clock crystals, etc.  Going to 66 MHZ with a DX2 or 100 MHz would be much more likely.

75 MHz was often slower in practice than a 66 MHz machine and was really only ever used to upgrade a fixed-frequency 486-25 when you didn't know how to change a clock crystal or the rest of the chipset was so old that it only supported 25 MHz FSB.

Quote
Yeah, most only needed one jumper so the other is usually in a 'storage spot'

Yeah, there are always places to stick a spare jumper so it will have no effect.  The Maxtor recommended way for those model drives is illustrated above.  This differs on other makes or other models of Maxtor, though.

Quote
For the yougins....

You set master/slave by setting the jumpers, IIRC windows liked to be the C: drive, so windows drive master second or cdrom etc slave. Or if it was ATA66 (IIRC) and you had a compatible cable (master connection = master drive, slave = slave drive) CS on the jumpers.

I don't know what you mean...  Drive letters really have nothing to do with what the physical drive arrangement is.  Windows has never cared what drive letter you installed it on and back in the day, and back on early versions of Windows, when you had the 32 meg partition limit (especially in the MFM/RLL days,) without using something like OnTrack Disk Manager to get larger partitions, with a 40, or especially 80+ meg drive you already had at least a drive D:, E: or F:, never mind the letters on multiple drive systems.  I usually installed Windows 286 on the D: or E: partition on my machines, for example.

Quote
Also, jumpers were required for things like setting the clock/voltage of the CPU.

All this new fangled automagic bios stuffs now.

It goes back a lot farther than clock/voltage.  :)  On the original PC you had to set things like disk drives and memory configuration.  Before that it was things like terminal baud rate or line printer interface card I/O address.   ;)
« Last Edit: February 12, 2018, 07:00:06 am by drussell »
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1116
  • Country: ca
  • Hardcore Geek
Re: OLD COMPUTER
« Reply #14 on: February 12, 2018, 06:52:57 am »
Also, in terms of upgrade, unless an overdrive chip is found, it's possible that machine can only go up to a 50 mhz upgrade (normally 40mhz if you have VLB cards). 486es, though, fun machines to work with.

Uh, no...  a 50 MHz DX2 would be running with a 25 MHz bus, so the VLB (on a later machine, that had VLB, which the OP's machine almost certainly does not since it is only 33 MHz, unless they're using a late model board with a very early chip) would be fine with a 50 or 100 MHz DX2/DX4.  VLB was "supposed" to run at 33 MHz, and is what most of the cards were specced for.  It was the later 40 MHz bus clocks that caused problems with some cards, especially if you didn't know how to set everything up properly.  I have several AMD DX4-120s that I used to run at 160 MHz (40 MHz x 4), with Cirrus Logic VLB video (5428?) and VLB IDE controllers, often using dual hard drives in striped arrays to double the speed.  (This would be, like, about 420 meg drives probably.)   

I think I even still have a couple of Adaptec VLB SCSI controllers around here somewhere.  I can't remember if those worked fine at 40 MHz or not, but I think they usually did.  That was the hot hot set-up.  A couple Quantum 420 or 540 meg SCSI drives on a VLB Adaptec?  16 or 32 megs of RAM?  That's a $7000 workstation if it's got SCO on it or something.  I sold a few of those to an engineering firm when I was in high school.  They LOVED doing CAD on those Cirrus Logic cards.
« Last Edit: February 12, 2018, 06:58:20 am by drussell »
 

Offline Ampera

  • Super Contributor
  • ***
  • Posts: 2508
  • Country: us
    • Ampera's Forums
Re: OLD COMPUTER
« Reply #15 on: February 12, 2018, 01:08:07 pm »
I am actually wrong on one count, OverDrive chips can be used to upgrade those machines, not sure what I was thinking. However, what I was thinking, was clock multiplication was not something that could be done on boards without the jumpers for it without an OverDrive chip, and I still believe this to be the case.

So this is why I said 50mhz, as that is the fastest non-clock multiplied 486 chip ever released (The DX-50). I also think you have the DX-50 and the DX-40 and it's multiplications somehow mixed up, as the VLB specification perfectly supports 40mhz bus widths, and so does every single VLB card I own, which is two. It's the 50 mhz speed that is out of spec (according to Wikipedia).
C/C++/Java Programmer, Legacy hardware enthusiast, madman.
If it's broken, I probably did it.
EEVBlog IRC: irc.austnet.irc #eevblog
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1116
  • Country: ca
  • Hardcore Geek
Re: OLD COMPUTER
« Reply #16 on: February 12, 2018, 04:08:50 pm »
Ah, yes...  Come to think of it, it was probably 50 MHz x 3 I was running those AMD chips at...

I should pull a couple of them out and see what I was running!  :)
 

Offline imidis

  • Frequent Contributor
  • **
  • Posts: 427
  • Country: ca
Re: OLD COMPUTER
« Reply #17 on: February 12, 2018, 05:27:57 pm »
Eh, I never said my memory was perfect! Perhaps it was windows wanted the master  :-// Too long ago I guess
Gone for good
 

Offline Naguissa

  • Regular Contributor
  • *
  • Posts: 108
  • Country: es
    • Foro de electricidad, electrónica y DIY / HUM en español
Re: OLD COMPUTER
« Reply #18 on: February 12, 2018, 05:57:28 pm »
I am actually wrong on one count, OverDrive chips can be used to upgrade those machines, not sure what I was thinking. However, what I was thinking, was clock multiplication was not something that could be done on boards without the jumpers for it without an OverDrive chip, and I still believe this to be the case.

So this is why I said 50mhz, as that is the fastest non-clock multiplied 486 chip ever released (The DX-50). I also think you have the DX-50 and the DX-40 and it's multiplications somehow mixed up, as the VLB specification perfectly supports 40mhz bus widths, and so does every single VLB card I own, which is two. It's the 50 mhz speed that is out of spec (according to Wikipedia).
I had a 486-dx 50 (no multiplier) with a Cirrus Logic VLB and it worked ok, so maybe out of spec, but working


Edit: Intel dx, not dx2.


Enviado desde mi Jolla mediante Tapatalk
« Last Edit: February 12, 2018, 05:59:25 pm by Naguissa »
 

Offline Ampera

  • Super Contributor
  • ***
  • Posts: 2508
  • Country: us
    • Ampera's Forums
Re: OLD COMPUTER
« Reply #19 on: February 12, 2018, 06:51:49 pm »
There are cards that will work, but you are effectively overclocking them.
50x3 would definitely be a sight to behold. Sadly my 486 board only goes up to 40x3 which is what I have my DX4-100 at (yes I've overclocked it!)

Either way, for this machine, I believe it's the OverDrive chips that allowed for clock multiplication without motherboard support (they did the clock doubling on-chip) which is why they were so special. They also had onboard VRMs, so if you can find a 486 OverDrive (or if you have Socket 2, maybe splurge on a Pentium OverDrive) you can try those out.
C/C++/Java Programmer, Legacy hardware enthusiast, madman.
If it's broken, I probably did it.
EEVBlog IRC: irc.austnet.irc #eevblog
 

Online james_s

  • Super Contributor
  • ***
  • Posts: 9579
  • Country: us
Re: OLD COMPUTER
« Reply #20 on: February 12, 2018, 08:38:12 pm »
I vaguely remember the DX-50 and reading that it was faster than the DX2-66 although back then I didn't really understand why. I seem to recall that the DX-50 systems had a reputation for being flaky, 50MHz FSB was really bleeding edge tech at the time. I don't think I ever saw one in the flesh, I remember my friend's dad got a 486-33 that seemed screaming fast, I eventually upgraded to a 486DLC-40 which IIRC was a low power CPU designed for laptops that ran on a 386DX motherboard. I went from that to a Pentium-60, huge hot running 5V ceramic/gold package, it blew me away the first time I fired up Doom on that thing.
 

Offline drussell

  • Super Contributor
  • ***
  • Posts: 1116
  • Country: ca
  • Hardcore Geek
Re: OLD COMPUTER
« Reply #21 on: February 12, 2018, 10:46:24 pm »
...
I eventually upgraded to a 486DLC-40 which IIRC was a low power CPU designed for laptops that ran on a 386DX motherboard.

Yeah, the Cyrix 486DLC was a decent chip.  I used one in my BBS machine and then, I think, my first FreeBSD machine but then because I didn't have a 387 that could run at 40 MHz (I think I had a -16 that would run to 20 MHz) and the software math emulation caused all sorts of issues in the FreeBSD 1.x days, I ended up going and buying my first 486 for myself to make into the UNIX box.  :)  Ahh...  The good ol' days.

The 486DLC was a chip that essentially supported all of the 486 instruction set but was packaged as a 386 and ran with a normal 386 chipset.
 

Offline Halcyon

  • Moderator
  • *****
  • Posts: 3684
  • Country: au
Re: OLD COMPUTER
« Reply #22 on: February 12, 2018, 11:30:21 pm »
Yeah, the Cyrix 486DLC was a decent chip.
The 486DLC was a chip that essentially supported all of the 486 instruction set but was packaged as a 386 and ran with a normal 386 chipset.

They were? My memories of Cyrix CPU's were that they were relatively slow (compared with the Intel and AMD counterparts). I could be mistaken though, I only ever owned 2 machines with Cyrix CPUs in them.

I will say they were more stable and reliable compared to AMD CPU's of the era.
 

Online james_s

  • Super Contributor
  • ***
  • Posts: 9579
  • Country: us
Re: OLD COMPUTER
« Reply #23 on: February 13, 2018, 06:59:44 am »
The Pentium class Cyrix chips were slow, largely due to their FPU design as I recall. Prior to that era FPUs were mostly only used with certain professional software but then 3D games appeared that relied heavily on floating point calculations. Cyrix had some innovative stuff in the 386/486 era, parts like the DLC were a significant upgrade to a 386 machine.

It's pretty much always been the case that CPU companies occasionally leapfrog each other.
 

Offline Ampera

  • Super Contributor
  • ***
  • Posts: 2508
  • Country: us
    • Ampera's Forums
Re: OLD COMPUTER
« Reply #24 on: February 13, 2018, 02:42:05 pm »
Ever since 2011, to this date, the CPU industry has gotten interesting, with AMD's effective dropping off the face of the performance charts, when Intel made their platform a bit pointless with Sandy Bridge.
Now that AMD is back in the game, the performance balancing act is going to get interesting, especially with the looming event of the end of Moore's law, where CPU manufacturers will have to start to get more creative.

I've never actually owned a Cyrix chip, and to my knowledge I have never owned an x86 CPU not made by AMD or Intel (I might have and just not know it in some sort of embedded application).
C/C++/Java Programmer, Legacy hardware enthusiast, madman.
If it's broken, I probably did it.
EEVBlog IRC: irc.austnet.irc #eevblog
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf