Thx, how do I install the XTIDE-Bios in the IDE Slot? Have you done this before?

Uhm, which IDE slot first of all?

dr.zeissler wrote on 2021-09-14, 20:09:

Thx, how do I install the XTIDE-Bios in the IDE Slot? Have you done this before?

I while ago i saw a XTIDE BIOS ISA Card on https://www.tindie.com/. But i could not find it now. I basically makes your existing IDE controller compatible with newer hard drives.
You can use an ordinary XTIDE ISA Card with IDE or CF interface. They will also replace the built in HDD support in your BIOS with the newer XTIDE BIOS. But it will use the ISA bus.
SOm Network cards also have a socket for BIOS - you can get an XTIDE Bios and put it on the Network card.

Have not tried this.

konc wrote on 2021-09-15, 06:45:

Uhm, which IDE slot first of all?

Sorry i meant the ISA slot.

ah, ok!

Jinxter wrote on 2021-09-15, 19:32:

konc wrote on 2021-09-15, 06:45:

Uhm, which IDE slot first of all?

Sorry i meant the ISA slot.

Ah alright, this makes sense!

It wasn't exactly wrong, though.
Long ago, IDE HDDs were also called AT-Bus HDDs/Fixed Disks.
That's where ATAPI and ATA comes from (AT Attached).
IDE is a sub set of the PC/AT-Bus (now commonly being referred to as ISA).

Jo22 wrote on 2021-09-16, 12:51:

It wasn't exactly wrong, though. Long ago, IDE HDDs were also called AT-Bus HDDs/Fixed Disks. That's where ATAPI and ATA comes f […]
Show full quote

It wasn't exactly wrong, though.
Long ago, IDE HDDs were also called AT-Bus HDDs/Fixed Disks.
That's where ATAPI and ATA comes from (AT Attached).
IDE is a sub set of the PC/AT-Bus (now commonly being referred to as ISA).

This is an XT though 😉

konc wrote on 2021-09-16, 14:38:

Jo22 wrote on 2021-09-16, 12:51:

It wasn't exactly wrong, though. Long ago, IDE HDDs were also called AT-Bus HDDs/Fixed Disks. That's where ATAPI and ATA comes f […]
Show full quote

It wasn't exactly wrong, though.
Long ago, IDE HDDs were also called AT-Bus HDDs/Fixed Disks.
That's where ATAPI and ATA comes from (AT Attached).
IDE is a sub set of the PC/AT-Bus (now commonly being referred to as ISA).

This is an XT though 😉

Oh. Right. My bad.😅

Then let's pretend it has an 8086/V30 CPU and a full 16-Bit data path. Then it's almost ISA. Sorta.. 😁👍

Some XT compatibles had pseudo-ISA with 16-Bit I/O, but merely 20-Bit address space.
Minus the upper IRQ lines and the second DMA controller.
For example, the Olivetti M24, S100 systems or Sergey's XT replica.

https://www.retrospace.net/infoseiten/readm.php?id=48

http://www.s100computers.com/My%20System%20Pa … CPU%20Board.htm

http://www.malinov.com/Home/sergeys-projects/sergey-s-xt

But yeah, you're right. A true ISA system uses an 80286 or higher CPU.

What is "true" ISA?

It is either PC-Bus or AT-Bus ISA. https://en.wikipedia.org/wiki/Industry_Standard_Architecture

Caluser2000 wrote on 2021-09-17, 00:40:

What is "true" ISA?

It is either PC-Bus or AT-Bus ISA. https://en.wikipedia.org/wiki/Industry_Standard_Architecture

Neither?

After the 32-Bit EISA bus was defined by the Gang of Nine,
the specs for the previous 16/24-Bit AT bus were set in stone as a side effect.
That was necessary, because EISA had the old slot contacts as a legacy built-in.
The resulting "ISA" bus was similar to the AT bus, but not exactly the same.
ISA describes the timings for the bus, including a reference frequency of 8.33 MHz.

AT bus differed from this. The original PC AT Model 5170 ran at 6 MHz, originally, and had no upper limit yet (open ended, if you will).
(That's why some late 80s expansion boards were specced for up to 12MHz.)

According to the later ISA specs, that 6MHz PC AT must have run its bus out-of-spec, thus - if it was considered an implementation of ISA, which it wasn't.
To be fair, though, later ATs increased operation frequency to 8MHz, which still was a bit below 8.33 MHz.

However, that's not all, bus cycles might have been different, not sure.
You know, rising and falling edges. That low-level stuff.
To be honest, I'm too lazy to check right now.

Anyway, these are merely details.
It's comparable to AT-Bus vs IDE vs ATA0, I believe.
Most people will make no distinction between these.

To them, AT bus equals ISA equals and PC bus equals XT bus equals 8-Bit "ISA".
Which in most cases is close enough.
But there are subtle differences sometimes.
The original PC 5150 had one slot that was special, I vaguely remember.

Anyway, I just think that's interesting. 🙂

Jo22 wrote on 2021-09-17, 19:33:

Neither? […]
Show full quote

Caluser2000 wrote on 2021-09-17, 00:40:

What is "true" ISA?

It is either PC-Bus or AT-Bus ISA. https://en.wikipedia.org/wiki/Industry_Standard_Architecture

Neither?

After the 32-Bit EISA bus was defined by the Gang of Nine,
the specs for the previous 16/24-Bit AT bus were set in stone as a side effect.
That was necessary, because EISA had the old slot contacts as a legacy built-in.
The resulting "ISA" bus was similar to the AT bus, but not exactly the same.
ISA describes the timings for the bus, including a reference frequency of 8.33 MHz.

AT bus differed from this. The original PC AT Model 5170 ran at 6 MHz, originally, and had no upper limit yet (open ended, if you will).
(That's why some late 80s expansion boards were specced for up to 12MHz.)

According to the later ISA specs, that 6MHz PC AT must have run its bus out-of-spec, thus - if it was considered an implementation of ISA, which it wasn't.
To be fair, though, later ATs increased operation frequency to 8MHz, which still was a bit below 8.33 MHz.

However, that's not all, bus cycles might have been different, not sure.
You know, rising and falling edges. That low-level stuff.
To be honest, I'm too lazy to check right now.

Anyway, these are merely details.
It's comparable to AT-Bus vs IDE vs ATA0, I believe.
Most people will make no distinction between these.

To them, AT bus equals ISA equals and PC bus equals XT bus equals 8-Bit "ISA".
Which in most cases is close enough.
But there are subtle differences sometimes.
The original PC 5150 had one slot that was special, I vaguely remember.

Anyway, I just think that's interesting. 🙂

Edit: Waiting for the first edit...😉

^Not after being quoted, that would be against the netiquette. 😀

Jo22 wrote on 2021-09-17, 19:53:

^Not after being quoted, that would be against the netiquette. 😀

That is the whole idea.

Jo22 wrote on 2021-09-17, 19:33:

The original PC 5150 had one slot that was special, I vaguely remember.

No, the special slot is not in the 5150, but in the 5160. It's "slot 8" which is driven by a different data path, and thus needed a "CARD SELECT" signal to activate that data path. I suspect the fan-out of the data drivers for the XT bus was insufficient for 8 cards, so IBM had to add a dedicated buffer for the eigthth slot. To avoid bus conflicts on reads, the CARD SELECT signal on slot 8 toggled whether the main XT-bus data buffer or the slot-8 data buffer was used to drive the data from the card to the processor bus. The 5150 had only 5 slots, so the issue wasn't present in that machine.

CARD SELECT was so obscure that the same pin got repurposed to 0WS on the AT bus.

mkarcher wrote on 2021-09-17, 21:38:

Jo22 wrote on 2021-09-17, 19:33:

The original PC 5150 had one slot that was special, I vaguely remember.

No, the special slot is not in the 5150, but in the 5160. It's "slot 8" which is driven by a different data path, and thus needed a "CARD SELECT" signal to activate that data path. I suspect the fan-out of the data drivers for the XT bus was insufficient for 8 cards, so IBM had to add a dedicated buffer for the eigthth slot. To avoid bus conflicts on reads, the CARD SELECT signal on slot 8 toggled whether the main XT-bus data buffer or the slot-8 data buffer was used to drive the data from the card to the processor bus. The 5150 had only 5 slots, so the issue wasn't present in that machine.

CARD SELECT was so obscure that the same pin got repurposed to 0WS on the AT bus.

Well caught Dude.

mkarcher wrote on 2021-09-17, 21:38:

Jo22 wrote on 2021-09-17, 19:33:

The original PC 5150 had one slot that was special, I vaguely remember.

No, the special slot is not in the 5150, but in the 5160. It's "slot 8" which is driven by a different data path, and thus needed a "CARD SELECT" signal to activate that data path. I suspect the fan-out of the data drivers for the XT bus was insufficient for 8 cards, so IBM had to add a dedicated buffer for the eigthth slot. To avoid bus conflicts on reads, the CARD SELECT signal on slot 8 toggled whether the main XT-bus data buffer or the slot-8 data buffer was used to drive the data from the card to the processor bus. The 5150 had only 5 slots, so the issue wasn't present in that machine.

CARD SELECT was so obscure that the same pin got repurposed to 0WS on the AT bus.

Ah, I see. Thank you very, very much for taking your time for the detailed explanation! 😎👍
- I merely remembered it was somehow related to the original PC models (PC, XT).

It's things like this why I love this place so much, by the way.
In other forums, someone would likely just say "you're wrong" without giving a hint why that's the case.

I have just found myself an EuroPC II. It has 768KB of RAM.
Does anybody know hos to use this 128KB of extra EMS? - The startup ram checker counts it as EMS.

The BIOS in Euro PC say mentions JIM (v1) and Sara (V0) in IO 250h and 25Eh - What does this mean?

Jinxter wrote on 2021-10-17, 18:19:

I have just found myself an EuroPC II. It has 768KB of RAM.
Does anybody know hos to use this 128KB of extra EMS? - The startup ram checker counts it as EMS.

I think USE!UMBS and DOSMAX might suit you here. USE!UMBS to enable the high mem and dosmax to move command.com and others to high mem to free up conventional mem. Clearmem may be needed to clear memory blocks first.