weedeewee wrote on 2022-05-26, 07:30:

FreddyV wrote on 2022-05-26, 06:06:

The ROM on this board can t be used as there is a bug in @ decoding.

Is the rom not being able to be used, a hardware problem or software?

It is an Hardware problem, some Address lines are not connected.
It is corrected in the LoTech design of this board.

Jinxter wrote on 2022-05-27, 10:25:

FreddyV wrote on 2022-05-25, 17:05:

So here is what I did with the Driver 😀 https://youtu.be/WNg8ROz0weg

Wow, really nice work. Thank you.

Will you be working on making a boot rom?

A Boot ROM already exist, it has some problem apparently but I am not able to test it for the moment.
I confirm anyway that its code is really close to this Driver, written with the same base functions, and as slow as the driver for 8086/286

I need to learn more about the DISK BIOS Code, mainly on the way to store data in memory, the Disk geometry, the Previous Disk BIOS code, variables and so on.
This BIOS copy itself on the top of the memory and modify the RAM Size detected by the BIOS, but I am not sure it is really what we need to do.
I need help from an XT IDE Code Expert.

Have a BOOT ROM will be really interesting if we are able to produce cheap boards without the @ Bug.

FreddyV wrote on 2022-05-27, 16:03:

It is an Hardware problem, some Address lines are not connected.
It is corrected in the LoTech design of this board.

Most of us Retro fans are able to use a soldering iron. Is the bug anything we can fix ourself?

Jinxter wrote on 2022-05-27, 11:16:

Here is a USBISA card bracket. I am using it. Works fine.
https://www.thingiverse.com/thing:5350142

That looks familiar 😀 Glad to see that folks are using it!

FreddyV wrote on 2022-05-27, 16:03:

weedeewee wrote on 2022-05-26, 07:30:

FreddyV wrote on 2022-05-26, 06:06:

The ROM on this board can t be used as there is a bug in @ decoding.

Is the rom not being able to be used, a hardware problem or software?

It is an Hardware problem, some Address lines are not connected.
It is corrected in the LoTech design of this board.

Well now, that's just Frickin' silly.
Could you fix it with some insulated wire? 😀

weedeewee wrote on 2022-05-27, 17:12:

Well now, that's just Frickin' silly.
Could you fix it with some insulated wire? 😀

No, you need to add a component

Hi,

New version 😀

I did multiple version for different CPU, it is also faster for Write.

(I removed the 2.0 bugged version)

Btw calamitylime is working on a pcb based bracket that you guys could use USB to Serial Mouse using the Pi Pico! it’s buried in this thread

Also, what does this driver do again? Make transfer faster? Or add features?

Hi,

I found a bug in V0.1 and V0.2 that may generate corruption if a software set the DIR Flag...

Here is the "Bug" fix

Sphere478 wrote on 2022-05-29, 18:59:

Btw calamitylime is working on a pcb based bracket that you guys could use USB to Serial Mouse using the Pi Pico! it’s buried in this thread

Also, what does this driver do again? Make transfer faster? Or add features?

Hi,

It is faster ( 10x) than the original driver on 8086/8088 and 286.
Also it set the Port 260 and Speed 0 by default even if there is no parameter.

There is no fix for the moment (Except for the bug I added 😀 ), as no bug was reported so far.

I saw that some folder are corrupted in my DOS 3.31 Image (Under DOS 3.31) so I may take a look at this.

I also saw it support only 512b Sectors, I will correct this. (But is it really needed ?)

Will a driver (different one) be able to turn this into a actual usb port?

Sphere478 wrote on 2022-05-29, 21:10:

Will a driver (different one) be able to turn this into a actual usb port?

I think people are just trying to get the storage driver working well right now.

While it seems conceivable that someone might make an input driver for a mouse or keyboard, I imagine if anyone ever writes such a driver, it would likely be a separate driver that would not be loaded at the same time as the storage driver initially.

Hubs and multiple devices would add increasing levels of complexity and driver memory consumption. Might be theoretically possible. Doesn't seem practicable in the next few years.

douglar wrote on 2022-05-29, 23:38:

I think people are just trying to get the storage driver working well right now. […]
Show full quote

Sphere478 wrote on 2022-05-29, 21:10:

Will a driver (different one) be able to turn this into a actual usb port?

I think people are just trying to get the storage driver working well right now.

While it seems conceivable that someone might make an input driver for a mouse or keyboard, I imagine if anyone ever writes such a driver, it would likely be a separate driver that would not be loaded at the same time as the storage driver initially.

Hubs and multiple devices would add increasing levels of complexity and driver memory consumption. Might be theoretically possible. Doesn't seem practicable in the next few years.

I was thinking for 9x or xp.

Maybe a existing usb driver could be modified to work with this for that purpose.

I’ve been using internal hubs off the onboard usb. So not super critical, but would be cool.

I usually have a bunch of stuff taking up the pci and not much in the isa. So a usb isa would be handy

Sphere478 wrote on 2022-05-30, 02:23:

I was thinking for 9x or xp. […]
Show full quote

I was thinking for 9x or xp.

Maybe a existing usb driver could be modified to work with this for that purpose.

I’ve been using internal hubs off the onboard usb. So not super critical, but would be cool.

I usually have a bunch of stuff taking up the pci and not much in the isa. So a usb isa would be handy

For 9x and XP is is a huge effort for machines having PS/2 and for most of them already an USB Port.

These boards are definitively better for machines with an ISA BUS Only, so < Pentium

And Yes, it is possible to put something else on these boards, and it need a driver.

To be really usable with a Mouse, we will need 2 USB Port or support of USB HUB as use it as a HDD is really nice.

I received some test result for the Driver V2.01 using the lotech board.

IBM 5160 : 8088 4.77MHz
V88 : Write = 120.08 Kb/s | Read = 141.49 Kb/s
V86 : Write =127.24 Kb/s | Read = 121.87 Kb/s
V16 : Write = 147.13 Kb/s | Read = 170.24 Kb/s

On a 8088 8MHz, it reach 200Kb/s in Read (Bit Mode)

like I was saying, for me it would be handy because I usually have the pci populated with other cards. a usb isa card would be super handy.

those numbers sound pretty good for as slow as those cpus are. the numbers also suggest the cpu may be the limiting factor? though I suppose better drivers help with that.

good job 😀

Sphere478 wrote on 2022-05-30, 10:22:

like I was saying, for me it would be handy because I usually have the pci populated with other cards. a usb isa card would be super handy.

Have you looked into getting a good quality usb hub?

douglar wrote on 2022-05-30, 23:33:

Sphere478 wrote on 2022-05-30, 10:22:

like I was saying, for me it would be handy because I usually have the pci populated with other cards. a usb isa card would be super handy.

Have you looked into getting a good quality usb hub?

That’s what I did. It’s fine. A internal one.

Some boards don’t have usb though. So not always an option

Adding More USB Ports to Retro Motherboards

Did anybody test my driver ?

I worked on the 286 Code last Week, and read is around 320Kb/s on a 286 8MHz

FreddyV wrote on 2022-06-17, 20:01:

Did anybody test my driver ?

Still waiting on the card to arrive ….