Hey there everyone!
After some delays, I got modified versions of the cutemouse driver working.
I recompiled CuteMouse version 1.9.1, 2.0 and 2.1, just changing the latch divider parameter to work at different baud rates.
There's support for 1200, 2400, 4800, 9600, 19200 Baud in every language which cutemouse is available in.
The compiled driver's are available from the GitHub!
https://github.com/LimeProgramming/USB-serial-mouse-adapter
-Lime
Asus P/I P55T2P4(S,D), one of the Best HX-Boards, going 66 to 75 to 83(!!) Mhz, does exist in various Versions, and many are not able to go to the 512MB. From Version 3 on, they do, and if you've got the "right" Board you get the Voltage down to 2,0V
, so feel free to put on that Board an 400 AMD K6III+@75FSB=450 and be happy
Greetings, duncan
Ahh i see.
In that case yes the mod I linked should bring those boards up to 512.
It's not the easiest of soldering and the HX chip is very tthin but it should work just fine.
I've also been working on a mod for the VRM on my only socket 7 motherboard that should bring the core voltage down to 1.6v and maybe even run the vcore off of the 12v rail of the psu which should be cool.
-lime
....do I see that right, that this is a solution for all these poor sods with T2P4 1.x and 2.x boards, looking keen to our 3.x ?!?
I'm not familiar with an "t2p4"
Still new to this and I really only know the little I've come across.
Would you mind explaining what you mean?
Hello!
So I wrote up a guide about the dual tag ram mod for 430hx based motherboards.
I wrote the guide on vogons so here's the link:
https://www.vogons.org/viewtopic.php?f=46&t=88899
Hope you like it!
-Lime
Thinking about it, did i use percussive maintainance to solve an IRQ conflict?
Hello there!
Sorry for being a bit slow on this project.
I've been preoccupied with building and modding a socket 7 rig. Ye probably saw the image above with the two tag ram chips.
It's been quite the adventure.
Just mentioning this because I done a strange repair to a strange problem that others may find interesting.
Long story short:
I was having issues with using windows accessing the floppy drives and the serial mouse adapter at the same time. Every time the system would hang. I investigated IRQ conflicts but nope. Going as far as to use a multi io isa card for the floppy connection didn't work. Something was going on with the IRQ but I couldn't figure out what or even why it seemed to *sometimes* work. As in, 95% of the time, the system crashed with a complete dead stop but 5% of the time, it just spat out a random error.
It wasn't even a weird windows thing which I thought it was at first, it even happened under DOS.
So, I had a hunch that maybe the IRQ's are shorting somewhere on the board. DOS didn't care as much but windows did.
I had no idea where the issue would be but I thought the ISA slots would be a good place since there are dedicated IRQ pins going to every ISA slot. No short on those pins, so easy answer. I did note that the IRQ lines for the 8bit isa side are grouped separately to the 16 bits isa side, my issues were with the 8 bit ones so I followed them back to the multi-io chip on the edge of the motherboard. Checked continuity between the legs of the chip and the isa slot and everything checked out okay.I did note that each IRQ line was pulled high via a 6.5K resistor network.
Knowing that there's no way that the IRQ's go to that chip and stop there I checked the datasheet for an IRQ line out. The Lines that sounded promising didn't really lead me anywhere, so I followed out the IRQ lines.
Each IRQ of the chip went somewhere I could follow except IRQ_b as it was called. I traced it to the Parallel and IDE connector which didn't make any sense.
Eventually I figured out that IRQ_b was grounded and it looked like it went to the chipset.
I reflowed around the multi-io chip and no change, it stayed grounded. I saw nothing obvious and with the trace going in and around the ram slots I couldn't follow it, with it grounded I couldn't trace randomly hoping for the best.
Then I remembered that I had spotted and removed a flake of solder from under the chipset, what if I missed a bit?
SO I tapped the chipset with the ass of my screwdriver and the short went away.😶
I flushed out underneath the chipset with some compressed air and contact cleaner. Got a surprising amount of flux out from under the chip and the thing works now.
I can now use the serial mouse and my floppy drives! huzzah!
-Lime
Hello there
Well I'm not sure if this is the update people want but it's an update people are getting.
After tinkering with the source code for Cutemouse I believe I figured out how to make it operate at the various baud rates supported by the USB-2-232, I got the version 2.0 working @ 2400 baud so far. version 2.1 is a bit more complicated as it changed assemblers and said assembler has itself changed. I did manage to compile version 2.1 but the higher baud didn't work (but with the mess of files I had at that point I may have just compiled stock cutemouse by mistake) So I'll get to work on compiling cutemouse for the various baud rates. Right now it's one exe per baud rate and I'm not sure if I can make one exe to cover all the baud rates or not since it likes to default to mousesystems if all else fails but we'll see.
-Lime
Hello there
Sounds like a cool project, wish I could help but I know nothing of running websites on retro hardware.
It is at all secure even to do that for anything other than some fun?
-Lime
I know I should make a different thread for this but I'd just like to comment that this has been fun. It's just the right amount of smart and stupid that I crave more of in life.
I found the bug above while working on version 2.
Today I managed to get an experiment sorted for the KBD board. The serial terminal is now managed from Core 1 instead of Core 0 like before, the change is significant because Core 0 is required to be the core that writes to flash as well as manage TinyUSB. With core 1 now handling the terminal, core 0 is free to handle the converting USB keyboard commands to at keyboard commands.
This may not sound like much but it means that the adapter can be receiving and sending keyboards commands to your computer so your computer can send them back to the adapter over the serial port for the terminal!
Huzzah!
Version 1.3.2 released
https://github.com/LimeProgram…apter/releases/tag/v1.3.2
This is a bug fix release. The device would save your settings but not load them. I don't know how this got past my tests but it did.
Sorry about that.
I assembled and done some light testing to one unit.
It looks good and what limited functionality I tested, works well!
I am really happy to say that you can access the settings terminal over the usb mini-B port, even windows sets up the correct drivers for you!
This particular unit is going to be my dev unit.
So far there are only 2 errors in this board, the pwr led needs to be flipped and c12 need to be removed due to coil whine.
PCB's arrived today!
Still waiting for USB-2-232-KBD PCBs to arrive. The bits of LCSC have arrived at least.
Hopefully I'll have them before the month is out.
Wow you starting of, i had not even time to fully test the Mouse Adapter.
But it works fine and tracking speed is very good.
Even in 1600x1200 its useable
The hope is that the USB-2-232 will be rolled into the USB-2-232-KBD. So there is no wasted testing and messing around.
Improving one helps the other.
Handling the tracking on the usb-2-232 was quite the headache, one quick flick left or right gives enough movement to spin doom guy in a circle multiples times. I play modern games with that kind of sensitivity but it feels so wrong on older games.
A lot of effort went into making it feel transparent even though it cheats a little. Glad to hear that you find the tracking speed good. Did you get into the terminal okay in the end?
-Lime
Hallo!
It was long put off but I've finally ordered the PCB's from JLC with assembly. I ordered 10 since the price difference between 10 boards and 5 boards wasn't even a 10'er
I've checked and doubled checked and I'm pretty sure my circuit is correct for the parts I've ordered.
I've also ordered a tag ram modification board plus 2 tag ram chips for tinkering with that project!
-Lime
Hallo,
I decided to upload the exp slot cover and DS1287 Hat (renamed to DS128X Hat) to GitHub. Bare minimum upload and I could add pictures and stuff another time, just not in the zone for such things.
Three ds12885 dip chips arrived in the post today which is fun.
Also just to keep people here informed on something that's kind of exciting. I made a socket 7 motherboard repair thread on vogons and it seems to have shifted to hacking 2 tag ram chips to a HX motherboard that has only one tag ram chip on board and no SRAM module.
My own Socket 7 motherboard may wind up being a guinea pig for that. If that sounds like your thing, here's the link!
Maybe once there's a prototype I'll make a thread here for that.
Hallo!
So I've been trying to make the most out of ordering from JLC and I made two little boards to figure out panelizing.
One board is for mounting a PS/2 port on a serial port mount on my AT case.
The general idea is that the ps/2 port sticks to the PCB and the PCB is held to the case with a 3d printed bracket. Nothing special and panel mount ps/2 ports do exist but that's not as fun as making a PCB. I did order a handful of ps/2 ports from LCSC as part of my order with them for the USB-2-232-KBD
The second board is what I'm calling "DS1287 Hat"
I know scorp made the nwx287as a modern replacement but at least as far as I can see, the DS1685 and DS12885 in a full DIP package are cheaper than surface mount options.
The idea is that for powering a replacement XX85 chip, the user can bend up the pins for the crystal and battery to solder the pins to the board. OR, use the board as a hat for fixing the old fix where you dill a hole in the side of the old RTC. You could even sand down the top of the old RTC for clearance and keep the hat on for looks.
There's probably the same thing out there already but again, making my own board is more fun.
I plan to panelize them myself with KiKad and a handy tool but it can't combine milling the PCB's and vscoreing them. So that's why the notch to mark the top of the chip is a silkscreen instead of a cut, and JLC want the vscore lines on the edge cut layer so that's by the kicad render on the panel has a weird boarder.
A little annoying but eh, not a big deal.
I'm not sure I'll release the files for these, I don't think there would be any interest in them. I will have many many more than I need so I may give them out for free but you pay postage if someone wanted one.
-Lime
is it be able to use USB Mouse and USB Keyboard on same time ?
(USB Hub ?)
It doesn't have a hub built in but you can connect a hub or a KVM.
I actually don't know how much it would cost to build in a hub but hubs really don't cost that much generally.
Currently with the USB-2-232 it has some handling caked in for keyboards so it knows to ignore them but it is able to know it has both a keyboard and mouse plugged in. if it didn't then it would probably crash when connected to a KVM.
I should probably note that in the default TinyUSB lib, one hub does work but nested hubs do not. Which is why I had issues with KVM's. The modified version of TinyUSB is able to handle a chain of I think 10 4 port hubs, though I think I limited that for the sake of ram.
There we go.
Moved some traces around to keep vias away from the silk screen and to try to make them look nice.
The 5v power got itself a nice thicc trace, I don't think it needs it but I think it looks nice.
I also went ahead and renamed the project from USB-2-232-AT to USB-2-232-KBD
Hopefully it can be made support both AT and XT Keyboards, I haven't looked too deep into that topic and ultimately I lack the hardware to test it but to my understanding both standards are fairly similar with XT not having the same back and forth conversation that AT has.
-Lime
