Booting 720K in the 5150

[per]

Of course, that's a 16-bit multifunction card (just like the one in my 386) and won't work in XTs.

Using HD drives in an XT requires that you use an 8-bit floppy controller with onboard BIOS extension.

Unless you boot from a hard drive with a recent version of DOS. But you're right, you can't boot HD floppy disks with it unless you use a BIOS extension.

The original IBM controllers absolutely do work with 3.5" DD drives. The data rate is the same, and the late XTs were even available with them as a factory option.

I've acutally gotten one of those recently (havent arrived yet). However, didn't the XT only got 5.25" drive-bays? (I haven't seen a 5.25" wide 3.5" disk drive before...)

 

[Mike Chambers]

Of course, that's a 16-bit multifunction card (just like the one in my 386) and won't work in XTs.

Using HD drives in an XT requires that you use an 8-bit floppy controller with onboard BIOS extension.



The original IBM controllers absolutely do work with 3.5" DD drives. The data rate is the same, and the late XTs were even available with them as a factory option.



The BIOS in the late XTs recognizes 3.5" DD drives, but the earlier models don't know about anything except 5.25" DD, so you have to use DRVPARM or DRIVER.SYS to set it up properly.

i knew the data rate is the same on DD 5.25" and 3.5" but i didn't think the controller would access data past a certain point. you learn something every day!

but, about the 16-bit card... they absolutely work in an XT, the only component on it that takes advantage of the 16-bit extended pins is the IDE controller.

you can put that in any XT-class system and use it's serial, parallel, and floppy components.

 

[Fallo]

i knew the data rate is the same on DD 5.25" and 3.5" but i didn't think the controller would access data past a certain point. you learn something every day!

You probably thought that you couldn't use 80-track disks. In fact, that is partially true. If you have a pre-1986 BIOS (ie. before 3.5" disks were introduced to PCs) and you don't use DRVPARM to set DOS up properly, you can only access the first 40 tracks of a 3.5" disk and FORMAT won't let you format disks to 720k. If you try, you'll get a "Parameters not supported by drive" message.

but, about the 16-bit card... they absolutely work in an XT, the only component on it that takes advantage of the 16-bit extended pins is the IDE controller.

you can put that in any XT-class system and use it's serial, parallel, and floppy components.

Really? I'd think that the IDE controller would expect an IRQ signal that doesn't exist on 8-bit slots and lock up the whole system as a result. Unless maybe your card has a jumper to disable it.

 

[per]

Really? I'd think that the IDE controller would expect an IRQ signal that doesn't exist on 8-bit slots and lock up the whole system as a result. Unless maybe your card has a jumper to disable it.

IRQ's from IDE controllers are issued by the drives connected to the adapter. Since no drives are connected, no IRQ's will be requested or expected, and hence, those IRQ lines will not be used.

And if a drive was connected, and the IRQ was set to above 9, the computer wouldn't have noticed if an IRQ was requested, and therefore not able to respond (this will keep the IDE drive from keep requesting the IRQ, but it won't affect the rest of the bus).

 

[mikey99]

One of the more common mods for the PCjr back in the day
was adding a 3.5 / 720K external drive. I had mine attached as
the B drive, so I'm not sure if it had the ability to actually boot
from the 3.5.

 

[gerrydoire]

One of the more common mods for the PCjr back in the day
was adding a 3.5 / 720K external drive. I had mine attached as
the B drive, so I'm not sure if it had the ability to actually boot
from the 3.5.

When I had a XT back in the mid 80's, once the 720k Drive came out, I bought one, I used driveparm, at that time them small floppies seemed real cool!

 

[per]

When I had a XT back in the mid 80's, once the 720k Drive came out, I bought one, I used driveparm, at that time them small floppies seemed real cool!

Do you know if it was a 5.25" wide drive or a 3.5" wide drive in a 5.25">3.5" drive slot adapter?

 

[gerrydoire]

Do you know if it was a 5.25" wide drive or a 3.5" wide drive in a 5.25">3.5" drive slot adapter?

I think it was 5.25, not 3.5 in an adaptor, but I could be wrong, been so long.

I've been into the PC's since the XT days, I think it came with Dos 2.2.

Buying every new toy as it came out, payed a fortune for every inch of the way...

My XT came with Hercules and a Amber Monitor.

Then I bought ATI CGA, ATI EGA and on and on...

 

[Chuck(G)]

i knew the data rate is the same on DD 5.25" and 3.5" but i didn't think the controller would access data past a certain point. you learn something every day!

The core of legacy-interface floppy controllers has not changed from the earlies PC (and before)--it's still the Intel 8272/NEC uPD765 controller--basically a part made to work with the Intel 8080 CPU. Today, it may be buried in a corner of a support chip set, but it's still the same 30-year old design.

The BIOS in the late XTs recognizes 3.5" DD drives, but the earlier models don't know about anything except 5.25" DD, so you have to use DRVPARM or DRIVER.SYS to set it up properly.

Only if you'd like to format new media. Take, say, a DOS 3.0 boot diskette, hook a 3.5' drive to your 5150 and it'll boot, read and write just fine without driver changes. Formatting requires knowing the drive parameters, so a driver addition (either DRIVPARM or DRIVER.SYS) is needed.

 

[Great Hierophant]

I tried the 5.25 with a 360K floppy and it refused to work. I did not really test beyond setting the jumper to 300rpm instead of 360rpm. I was also using HD floppy disks with electrical tape over the disk type whole to fool the drive into treating the disks as DD disks. It seems that 720K DOS disks are not a problem for the 5150, but I believe that you should stick to 360K booters.

 

[Chuck(G)]

I tried the 5.25 with a 360K floppy and it refused to work. I did not really test beyond setting the jumper to 300rpm instead of 360rpm. I was also using HD floppy disks with electrical tape over the disk type whole to fool the drive into treating the disks as DD disks. It seems that 720K DOS disks are not a problem for the 5150, but I believe that you should stick to 360K booters.

Easy--5150 BIOS doesn't have the code to double-step the 5.25" drive for a 360K. Create a 720K boot on one, though and it'll work just fine.

 

[Mike Chambers]

You probably thought that you couldn't use 80-track disks. In fact, that is partially true. If you have a pre-1986 BIOS (ie. before 3.5" disks were introduced to PCs) and you don't use DRVPARM to set DOS up properly, you can only access the first 40 tracks of a 3.5" disk and FORMAT won't let you format disks to 720k. If you try, you'll get a "Parameters not supported by drive" message.



Really? I'd think that the IDE controller would expect an IRQ signal that doesn't exist on 8-bit slots and lock up the whole system as a result. Unless maybe your card has a jumper to disable it.

i didn't even bother disabling it on both of mine. since it's looking for a higher IRQ than what the machine can handle, the circuit isn't connected to an IRQ line. the system has no idea anything else besides the floppy/serial/parallel is on the card. no communication at all.

i might be wrong about this part, but i think that if the IRQ line isn't connected the IDE controller chip doesn't even activate. can somebody smart correct me if i'm wrong? trixter?

 

[Mike Chambers]

here, i booted up one of my 8088 machines (PC's Limited turbo board) that has a DLC 16-bit multi-i/o card in it. full 1.44 MB with it on an XT-class 8-bit ISA bus. one of my other 8088's is more of a true XT clone being only 4.77 MHz and built in 1984. it has a SIIG 16-bit card that works just the same.

 

[Chuck(G)]

i might be wrong about this part, but i think that if the IRQ line isn't connected the IDE controller chip doesn't even activate. can somebody smart correct me if i'm wrong? trixter?

Not smart, but experienced enough to know that the IRQ 14 line makes things a bit easier, but isn't required.

If desired, I can furnish a program to demonstrate that the primary IDE controller is still active on ports 1f1-1f7 and 3f6 no matter what's done with the IRQ line. A read of 1f7 clears an interrupt pending; that's why there's an alternate status register on 3f6.

I don't understand what you're trying to demonstrate, Mike. I don't think anyone's arguing that 8-bit peripherals can be accessed if plugged into an 8-bit ISA slot.