Of Mice and Serial Mice...

[TheManuel]

Background:
I have a CompUSA Crystal Trackball that will be used in my CP.
It is combination serial and PS/2.
It works great for a budget solution in PS/2 mode.

Problem:
When in serial mode, the movement resolution is obviously reduced and its performance simply sucks.  However, the absolutely most convenient alternative for me is to use it in serial mode to free up the PS/2 port for a spinner.

Question:
Is this a trait of all serial mice?
Is this a driver/software issue?
Does anyone know how to make it work well in serial mode?

[u_rebelscum]

...
Question:
Is this a trait of all serial mice?
Is this a driver/software issue?
Does anyone know how to make it work well in serial mode?

For the most part the PS/2 port is faster than the serial port (not including the really fast serials).  Since mice usually don't use the really fast serial ports, most of the time the serial port mice has a hardware speed limit under that of PS/2 mice.  Also, serial mice have a small, but bigger than PS/2, overhead in windows.

That said, check you serial port driver settings and see if you can increase the speed up to at least 4800-9600 kps (sometimes the default is 1200 kps).  The computer's serial I/O chips and the mouse need to be able to handle the higher speed for anything to be seen, though.  How to change the settings differers between dos, win95-98, winME, winNT, (and I'm not sure how on win2000 or winXP, so they might be different than the others too).

[AlanS17]

What version of Windows are you using??? I'm pretty sure XP lets you manually change your resolution (sample rate) in your mouse drivers.

[TheManuel]

Hi Robin.
You may soon want to start charging me for answering all of my qustions.  
Thanks for showing up again.

For the most part the PS/2 port is faster than the serial port (not including the really fast serials).  Since mice usually don't use the really fast serial ports, most of the time the serial port mice has a hardware speed limit under that of PS/2 mice.  Also, serial mice have a small, but bigger than PS/2, overhead in windows.

Considering I have a ghetto PC, it is pretty safe to say I have the crappier type of the serial ports.

That said, check you serial port driver settings and see if you can increase the speed up to at least 4800-9600 kps (sometimes the default is 1200 kps).  The computer's serial I/O chips and the mouse need to be able to handle the higher speed for anything to be seen, though.  How to change the settings differers between dos, win95-98, winME, winNT, (and I'm not sure how on win2000 or winXP, so they might be different than the others too).

Well, the COM1 setting in my computer (which is the serial port in this case) is set to the maximum of 128000 bits per second.  Changing this value to smaller ones, does not seem to have an effect on mouse behavior.
Here are other settings:
Data Bits = 8
Parity = none
Stop Bits = 1
Flow Control = none

By the way, its Windows XP.

Based on this, do you then think it's most likely a hardware limitation?
In that case, I will probably face the same issues with other serial mice.

Also check my answer to Alan's reply.

Thanks.

[TheManuel]

What version of Windows are you using??? I'm pretty sure XP lets you manually change your resolution (sample rate) in your mouse drivers.

It's Windows XP.
I checked if I could change those settings as you said.
It can be done for the PS/2 mouse but not for the serial.
It looks like I'm out of luck there.

Thanks .

[u_rebelscum]

Hi Robin.
You may soon want to start charging me for answering all of my qustions.
Thanks for showing up again.

You're welcome.  
I'm not helping fix this problem very much, though.  (see below)

...
Based on this, do you then think it's most likely a hardware limitation?
In that case, I will probably face the same issues with other serial mice.

Also check my answer to Alan's reply.

Thanks.

What version of Windows are you using??? I'm pretty sure XP lets you manually change your resolution (sample rate) in your mouse drivers.

It's Windows XP.
I checked if I could change those settings as you said.
It can be done for the PS/2 mouse but not for the serial.
It looks like I'm out of luck there.

Hmm, looks to me like either a driver or, more likely, a trackball hardward limit.  My guess is since you can't set the mouse refresh rate via the driver, either the driver sucks, or (more likely) the TB has a low fixed speed when plugged into the serial port.

I don't have one of those TBs, though.  Maybe someone with one can confirm that the serial is slower than mouse port on this specific device, or be of more help on it.  (As I posted before, though, serial mice are often slower than mouse port mice.)

You might be able to do better with a serial mouse to hack as the spinner.  However, a safer thing, especially with a winXP system, is do the spinner in USB instead of the mouse port, and leave the TB at it's better port.

[TheManuel]

You might be able to do better with a serial mouse to hack as the spinner.  However, a safer thing, especially with a winXP system, is do the spinner in USB instead of the mouse port, and leave the TB at it's better port.

There is another alternative which I discussed in this other post:
http://www.arcadecontrols.org/yabbse/index.php?board=1;action=display;threadid=7893

It's a long thread so I'll summarize it:
I'm basically considering using a spare PS/2 mouse I have
and wire its opto-electronics hack-style to the trackball PCB without disassembling the spare mouse in case I need it later for a direct connection to the PC.

As a result of discussions on that thread it looks like the only problem would be that the PC could get confused if it sees both PCB's.  I'm thinking maybe I can drill through the conductor tracks on the PCB's to isolate the diode and receiver and power them by direct wiring from the trackball PCB.  If I ever need the PCB working again, I can always re-fill the holes with soldering tin.

Does this sound like it's possible?

[_Iz-]

I'm thinking maybe I can drill through the conductor tracks on the PCB's to isolate the diode and receiver and power them by direct wiring from the trackball PCB.  If I ever need the PCB working again, I can always re-fill the holes with soldering tin.

Does this sound like it's possible?

It would be a bit of work but it should work. I think you'd have to wire power to both the emitter and the detector, ground to the emitter and 2 signal wires to the detector. Might be better to try and cut the traces with a razor knife instead of drilling them out.   Another option is to get an optic board somewhere (like oscar's or happ's universal optic boards) and use that.

[TheManuel]

It would be a bit of work but it should work. I think you'd have to wire power to both the emitter and the detector, ground to the emitter and 2 signal wires to the detector. Might be better to try and cut the traces with a razor knife instead of drilling them out.   Another option is to get an optic board somewhere (like oscar's or happ's universal optic boards) and use that.

The razor knife is a good idea.

The detector will need ground too, although your probably taking into account the on of the signal wires will be ground?

I'll look into oscar's boards.  Thanks for pointing that out.

[_Iz-]

The detector will need ground too, although your probably taking into account the on of the signal wires will be ground?

The detector only has 3 pins on it. Power and 2 signal. No (separate) ground required. Emitter has 2 pins, power and ground.

[TheManuel]

The detector only has 3 pins on it. Power and 2 signal. No (separate) ground required. Emitter has 2 pins, power and ground.

I just know very basic electronics but it seems to me that unless the emitter and detector are internally grounded together, one of the signal leads would be the signal itself and the other one the reference level (i.e. ground).  This ground would be essentially the same as the ground for the emitter.

I'll have to figure this out for the hack to work.  I'll buy a multimeter today. :-D