Monotech DeluxeFloppy giving excessive timeout

[Andrettigto]

Have a 5150 Rev.A with a Monotech DeluxeFloppy controller that's based on the ‎Sergey's Project. I was attracted to it because of its physical size and not its additional hardware coverage since I'm only using 5 1/4" DD and 3 1/2" DD drives. Currently I have it configured with drive A as the 5 1/4" since I have numerous old boot games.

The issue I'm having is after POST, there's a long delay (over 35 sec.) before timeout on an empty 5 1/4" drive, finally prompting for retry/fail input. Only then can I finally continue with XT-IDE booting to my CF or hard drive. With the standard IBM floppy controller, timeout only takes a few seconds.

Wondering if anyone can shed some light on this issue for me. I should note that with DeluxeFloppy configured to use the 3 1/2" as drive A, timeout is similar to the standard IBM controller occuring in less than 2 seconds.

 

[modem7]

... that's based on the ‎Sergey's Project.

So, the controller's BIOS will be a version of 'Sergeys Multi-Floppy BIOS'.

At [here], I documented some behaviour that I observed with version 2.2, and you can see that I too encountered an approximate 30 second delay.

 

[Trixter]

From https://github.com/skiselev/floppy_bios/blob/master/floppy_bios.asm:

fdc_motor_tout    equ    440h    ; byte - floppy drive motor off timeout (ticks)

Later, in https://github.com/skiselev/floppy_bios/blob/d6b2c204a2a9e8f40931a8b9053d57e4a5483d95/floppy1.inc:

	push	ds
	lds	si,[1Eh*4]		; DS:SI -> INT 1Eh
	mov	cl,byte [si+2]
	pop	ds
	mov	byte [fdc_motor_tout],cl ; ticks before turning off the motor

Past that, and who is setting 1Eh initially, I couldn't tell you.

 

[Andrettigto]

It does have version 2.2

Well commented code. So now, looking at it I'm wondering if the 720k drive has change-line support vs the 360k

 

[Trixter]

It does, and that's probably it. I almost mentioned it previously but forgot. DSDD 360K drives don't have change-line support. (That said, 30 seconds is excessive.)

 

[640KB]

To clarify, your 5150 Rev.A has the later 10/27/82 BIOS (otherwise option ROM controllers wouldn't be supported)? Once this times out eventually it will let you boot to the CF card - does this only happen after POST but before boot, and then the drive works normally in DOS after that? Does the same thing happen if you remove the XT-IDE CF adapter? Does it give any numerical error numbers?

Oh, the fdc_motor_timeout (40:40) is only used as the counter that turns off the drive motor after ~2 seconds of inactivity... not really related.

 

[Andrettigto]

Yes, 1501476 supporting adapters with ROM BIOS extensions. Once it times out, I can Fail the operation allowing XT-IDE boot menu and successfully selecting either my CF card or MFM drive. Afterwards, drive behavior is normal just like the OE IBM controller. The same 35+ seconds wait happened the last time I tried it in my other Rev.A that has no XT-IDE and only an MFM drive.

I'm finding that this card kind of spoils the experience when powering it on while someone's watching. As is, their boot times with parity checking is slow enough.

 

[640KB]

Yes, 1501476 supporting adapters with ROM BIOS extensions. Once it times out, I can Fail the operation allowing XT-IDE boot menu and successfully selecting either my CF card or MFM drive. Afterwards, drive behavior is normal just like the OE IBM controller. The same 35+ seconds wait happened the last time I tried it in my other Rev.A that has no XT-IDE and only an MFM drive.

Actually now that I think of it, I recall the same thing on my MultiFloppy as well with the 2.2 version. I believe this was changed in a later version of that ROM, since when I went to 2.6 it no longer had that behavior. There's a setting now for select which IPL code to use so that might be the difference. Do you have a way to flash a new image onto it and try 2.6?

I'm finding that this card kind of spoils the experience when powering it on while someone's watching. As is, their boot times with parity checking is slow enough.

I know, the 5150 memory check is brutal - though I believe it was actually sped up quite a bit on the 1501476 version. I have the 1981 ROM and with 512K I've timed it at about 1 minute, 30 seconds!

 

[Andrettigto]

Thanks, 640KB! Good to learn there's a later version that resolves this! I've been on the fence looking into purchasing a low cost programmer as I've always wanted to try doing it but seems like a steep learning curve. PROM-EPROM selections, 5150 pin adaptors... But to have the ability to upgrade or modify a BIOS, create diagnostic ROMs, no doubt it would be a really interesting and fun adventure.
Was looking into this low cost kit as a starting point.